I had been planning on building an o2 meter for my car for a while, but I finally got the motivation to actually do it when I found out I could use that as my final project for an Electrical Engineering course I was taking at UNH. The version that will go in my car will be soldered onto a propper project board and may actually only have one LED bank for space reasons, but the idea is the same. It should be noted (As Huw Powell pointed out to me) that the propper way for these lights to illuminate would be only 1 or 2 lights on in the "middle" of the display for roughly .45 volts (the stoich point of the oxygen sensor) and then grow outward to the left toward 0 volts for a lean condition and grow outward to the right toward 1 volt for a rich condition. It turns out that this would be somewhat difficult since the LM3914 features voltage regulation to the LEDs. The LM3914 can take almost any voltage you want over 3 volts, but what it does is fluctuate the voltage on the negative lead of the LED by about .4 volts to turn them on and off. Example; If Vcc = 12 volts, then when an LED is off, the + lead will have 12 volts and the negative lead will have 10.6. When the chip turns on an led, it does so by making v at the negative lead closer to 10.2 volts. This will create a potential difference of 1.8 volts across the LED which is sufficient to illuminate it (while 1.4 is not). Because of the difficulty of adapting this to work in a "growing outward" fashion, I'll probably just use one driver and 10 leds so that I can color code the level of danger (eg yellow for .3 volts, red for .1-.2, green for .4-.6, etc). Anyway, here is a really bad scan of what the breadboard looks like for this circuit. This is what happens when you need to get someting working ASAP for a grade ;-)

Here is a scan of the schematic that i did in EWB (Electronics Workbench)

You can get the EWB soft copy here. Modifications and ideas are welcome!

click here to go back to the images and info index.